Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

IBM Watson Health: How cognitive technologies have begun transforming clinical medicine and healthcare

Cite as: Kamel Boulos MN. IBM Watson Health: how cognitive technologies have begun transforming clinical medicine and healthcare (Oral session IV – Patient safety tools, Thursday 19 May 2016, 15:45-16:45, Hotel Puijonsarvi, Kuopio). In: Proceedings of the 4th Nordic Conference on Research in Patient Safety and Quality in Healthcare (NSQH2016), Kuopio, Finland, 18-20 May 2016 (organised by University of Eastern Finland), p.29. URL: http://www.uef.fi/NSQH2016 (In: Nykanen I (ed.). The 4th Nordic Conference on Research in Patient Safety and Quality in Healthcare. Kuopio, Finland, May 18-20, 2016. Program and Abstracts. Publications of the University of Eastern Finland. Report and Studies in Health Sciences 21. 2016, p.29 (of 119 p.). ISBN: 978-952-61-2130-7 (nid.), ISSNL: 1798-5722, ISSN: 1798-5730.)
IBM Watson health: how cognitive technologies have begun transforming clinical medicine and healthcare
Maged N Kamel Boulos
ABSTRACT
Background: IBM Watson Health (http://www.ibm.com/smarterplanet/us/en/ibmwatson/health/) belongs to a new generation of smart cognitive computing technologies (a type of artificial intelligence) that are poised to transform the way healthcare is delivered, and to vastly improve clinical outcomes, quality of care and patient safety.
Objectives: Our goal was to collect and document the huge potential of a range of emerging and exemplary uses of IBM Watson in healthcare in both developed and developing country settings.
Methods: A survey of current peer reviewed and grey literature has been conducted, looking for reports and case studies involving the use of IBM Watson in different health and healthcare applications.
Results, conclusions and clinical implications: With its ability to make sense of unstructured medical information by analysing the meaning and context of natural language, and uncovering important knowledge buried within large volumes of data and information, including medical images, IBM Watson is exceptionally well suited for clinical and healthcare decision support, where there are often elements of ambiguity and uncertainty. It has been (or is currently being) successfully deployed in many developed countries in the West, as well as in developing countries, such as India and South Africa. IBM Watson unlocks a complex case by acquiring information from multiple sources, e.g., accessing the electronic patient record, then parsing all related medical evidence at up to 60 million pages per second. After processing all of this information, Watson offers relevant and prioritised suggestions to the decision-maker, e.g., helping clinicians identify the best diagnosis and treatment options in complex oncology cases, and providing hospital managers with new operational insights. The ultimate goals are to reduce cost, medical errors, mortality rates, and help improve patients' quality of life.

  • Login to see the comments

IBM Watson Health: How cognitive technologies have begun transforming clinical medicine and healthcare

  1. 1. IBM Watson Health: How cognitive technologies have begun transforming clinical medicine and healthcare 19-20 MAY, KUOPIO, FINLAND Maged N. Kamel Boulos, MBBCh, PhD, SMIEEE UHI Professor and Chair of Digital Health, Scotland, UK maged.kamelboulos@uhi.ac.uk / mnkboulos@ieee.org t: @mnkboulos ♦ What is IBM Watson Health? ♦ Open to developers ♦ Applications & examples ♦ In developing countries ♦ Research evidence ♦ Alternative platform ♦ Conclusions
  2. 2. What is IBM Watson / Cognitive computing? • IBM Watson belongs to a class of AI (artificial intelligence) that (partially) simulates human thought processes and is known as 'cognitive computing'. • Cognitive computing includes self-learning systems that mimic to a great (but not full) extent the way the human brain works by harnessing data mining, pattern recognition and natural language processing. • Watson uses natural language capabilities, hypothesis generation, big data analysis and machine learning, and evidence-based learning and reasoning to support healthcare professionals as they make clinical decisions. The clinician starts by posing a query to the system, describing symptoms and other related factors. Watson then begins to parse the input to identify key pieces of information. The system supports medical terminology by design, extending Watson's natural language processing capabilities. • Watson next mines the patient data to find relevant facts about family history, current medications, other existing conditions, etc. It combines this information with current patient findings from tests and instruments, including medical images, which it can also "read", and then examines all available data sources to form hypotheses and test them. Watson can incorporate electronic medical record data, patient information, clinicians' notes, treatment guidelines, research, clinical studies and journal articles into the data available for analysis. • Watson will then provide a list of potential diagnoses along with a score that indicates the level of confidence for each hypothesis.
  3. 3. • "The ability to take context into account during the hypothesis generation and scoring phases of the processing pipeline allows Watson to address these complex problems, helping the doctor and patient make more informed and accurate decisions" (and avoid over-testing). http://www.ibm.com/innovation/ uk/watson/watson_in_healthcare .shtml • An enabler for precision medicine and P4 medicine. • Helps in designing better treatments.
  4. 4. Individual patient and population perspectives https://www.ibm.com/smarterplanet/us/en/ibmwatson/health/population https://www.ibm.com/smarterplanet/us/en/ibmwatson/health/solutions/patient- engagement
  5. 5. Not a replacement of clinicians, but a "cognitive prosthesis" for them
  6. 6. https://www.youtube.com/watch?v=cFUg-b9qeIk Video station
  7. 7. http://www.ibm.com/smarterplanet/us/en/ ibmwatson/watson-oncology.html
  8. 8. Open to developers https://www.ibm.com/smarterplanet/ us/en/ibmwatson/developercloud/ https://www.ibm.com/marketplace/cloud/simplify- crowd-sourced-research-studies/us/en-us http://www.ibm.com/smarterplanet/ us/en/ibmwatson/knowledge- studio.html
  9. 9. http://www.ibm.com/smarterplanet/us/en/ibmwatson/developercloud/services-catalog.html
  10. 10. http://www.ibm.com/smarterplanet/ us/en/ibmwatson/ecosystem.html Third-party health app examples http://theeliza.space/
  11. 11. http://www.medtronic.com/us-en/about-3/medtronic-ibm- watson-health.html Pathway Genomics OME™ app: https://www-03.ibm.com/press/us/en/pressrelease/48766.wss http://www-03.ibm.com/press/uk/en/pressrelease/49725.wss https://youtu.be/mULwISf9YCw https://nutrino.co/ and many more... 'Virtual Doctor' for diabetes Novo Nordisk A/S
  12. 12. Not just in the Western world, but also in developing countries https://www.manipalhospitals.com/IBM-Watson/ Video: https://youtu.be/hbqDknMc_Bo Bangalore, India
  13. 13. Not just in the Western world, but also in developing countries http://www.metropolitan.co.za/ South Africa Metropolitan Health is using IBM Watson Engagement Advisor Video: https://www.youtube.com/watch?v=vLE7VuppRzU
  14. 14. Peer-reviewed research evidence examples "Watson has been applied to a few pilot studies in the areas of drug target identification and drug repurposing. Results suggest that Watson can accelerate identification of novel drug candidates and novel drug targets by harnessing the potential of big data." http://dx.doi.org/10.1016/j.clinthera.2015.12.001
  15. 15. Peer-reviewed research evidence (Cont'd) IBM Watson in oncology http://dx.doi.org/10.1007/978-1-4939- 3283-2_10 (see pp.213-215)
  16. 16. Peer-reviewed research evidence (Cont'd) Reducing readmission rates: Using IBM Watson to automate outreach to selected high-risk patients can significantly reduce hospital readmission rates and healthcare costs http://dx.doi.org/10.1089/pop.2015.0014
  17. 17. Alternative platform: Google DeepMind Health https://deepmind.com/health.html • Google has been given access to an estimated 1.6 million NHS patient records from the Royal Free, Barnet and Chase Farm hospitals in London, going back over the past five years and continuing until 2017. The records include full names, as well as patient histories, but the data remain encrypted, meaning that Google employees should not be able to identify anyone, according to the Royal Free Trust. • Google DeepMind is a British AI start-up founded in 2010 and acquired by Google in 2014.
  18. 18. Conclusions • IBM started developing Watson in 2005. It reached a mature stage in 2011, heralding a new era of cognitive computing applications in many industries, including health and healthcare. • With its ability to make sense of unstructured medical information by analysing the meaning and context of natural language, and uncovering important knowledge buried within large volumes of data and information, including medical images and audio, IBM Watson is exceptionally well suited for clinical, healthcare and personal health decision support applications, where there are often elements of ambiguity and uncertainty. • Thanks to its openness to developers, it has been (or is currently being) successfully deployed, in hospital environments and in consumer apps, in many developed countries in the West, as well as in developing countries, such as India and South Africa.
  19. 19. Conclusions • As of May 2016, limited peer-reviewed literature exists about IBM Watson in health and medicine, and most available material is in the form of grey literature (thousands of news items, Web articles/reports and videos). However, more peer-reviewed research is expected in the near feature, as IBM Watson Health rapidly establishes itself as a key enabler technology for 'P5 medicine' (Precision, Predictive, Preventive, Personalised and Participatory) globally. • IBM Watson unlocks a complex case by acquiring information from multiple sources, e.g., accessing the electronic patient record, then parsing all related medical evidence at up to 60 million pages per second. After processing all of this information, Watson offers relevant and prioritised suggestions to the decision-maker, e.g., helping clinicians identify the best diagnosis and treatment options in complex oncology cases, and providing hospital managers with new operational insights (individual patient and population perspectives). • The ultimate goals of Watson-powered applications are to reduce healthcare costs, medical errors, morbidity and mortality rates, and help improve patients and populations' quality of life.
  20. 20. Appendix: Custom hardware • Both platforms, IBM Watson and Google DeepMind, employ their own proprietary specialised hardware to offer their cloud- based AI services to developers and users. Google's latest Tensor Processing Unit (TPU) https://cloudplatform.googleblog.com/2016/05/Google -supercharges-machine-learning-tasks-with-custom- chip.html IBM Watson uses a cluster of massively parallel POWER7 processors/servers

×